Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=inan20
Nanotoxicology
ISSN: 1743-5390 (Print) 1743-5404 (Online) Journal homepage: https://www.tandfonline.com/loi/inan20
Colonizing microbiota protect zebrafish larvae
against silver nanoparticle toxicity
Bregje W. Brinkmann, Bjørn E. V. Koch, Herman P. Spaink, Willie J. G. M.
Peijnenburg & Martina G. Vijver
To cite this article: Bregje W. Brinkmann, Bjørn E. V. Koch, Herman P. Spaink, Willie J. G. M. Peijnenburg & Martina G. Vijver (2020) Colonizing microbiota protect zebrafish larvae against silver nanoparticle toxicity, Nanotoxicology, 14:6, 725-739, DOI: 10.1080/17435390.2020.1755469 To link to this article: https://doi.org/10.1080/17435390.2020.1755469
© 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
View supplementary material
Published online: 23 Apr 2020.
Submit your article to this journal
Article views: 336
View related articles
ARTICLE
Colonizing microbiota protect zebrafish larvae against silver
nanoparticle toxicity
Bregje W. Brinkmanna, Bjørn E. V. Kochb, Herman P. Spainkb, Willie J. G. M. Peijnenburga,c and Martina G. Vijvera
a
Institute of Environmental Sciences (CML), Leiden University, Leiden, the Netherlands;bInstitute of Biology (IBL), Leiden University, Leiden, the Netherlands;cCenter for Safety of Substances and Products, National Institute of Public Health and the Environment (RIVM), Bilthoven, the Netherlands
ABSTRACT
Metal-based nanoparticles exhibiting antimicrobial activity are of emerging concern to human and environmental health. In addition to their direct adverse effects to plants and animals, indir-ect effindir-ects resulting from disruption of beneficial host–microbiota interactions may contribute to the toxicity of these particles. To explore this hypothesis, we compared the acute toxicity of silver and zinc oxide nanoparticles (nAg and nZnO) to zebrafish larvae that were either germ-free or colonized by microbiota. Over two days of exposure, germ-free zebrafish larvae were more sensi-tive to nAg than microbially colonized larvae, whereas silver ion toxicity did not differ between germ-free and colonized larvae. Using response addition modeling, we confirmed that the pro-tective effect of colonizing microbiota against nAg toxicity was particle-specific. Nearly all mortal-ity among germ-free larvae occurred within the first day of exposure. In contrast, mortalmortal-ity among colonized larvae increased gradually over both exposure days. Concurrent with this gradual increase in mortality was a marked reduction in the numbers of live host-associated microbes, suggesting that bactericidal effects of nAg on protective microbes resulted in increased mortality among colonized larvae over time. No difference in sensitivity between germ-free and colonized larvae was observed for nZnO, which dissolved rapidly in the exposure medium. At sublethal con-centrations, these particles moreover did not exert detectable bactericidal effects on larvae-associ-ated microbes. Altogether, our study shows the importance of taking host–microbe interactions into account in assessing toxic effects of nanoparticles to microbially colonized hosts, and pro-vides a method to screen for microbiota interference with nanomaterial toxicity.
ARTICLE HISTORY
Received 1 December 2019 Revised 27 March 2020 Accepted 2 April 2020
KEYWORDS
Fish embryo acute toxicity test; host–microbiota interactions; particle-specific toxicity; gnotobiotic techniques; germ-free
1. Introduction
Microbiota that reside in and on plants and animals, interact closely with their hosts, modulating immune responses, nutrient uptake, and energy metabolism (Hacquard et al.2015; Brugman et al. 2018). Healthy hosts with beneficial microbiota harbor diverse mutualistic and commensal microbes, yet restrict growth of pathogenic microbes. Perturbation of the interactions between hosts and interacting micro-biota, called ‘dysbiosis’, has been related to severe infections, metabolic disorders, and immune diseases across humans, animals, and plants (Willing, Russel, and Finlay2011). For this reason, the release of anti-microbial agents into the environment, potentially
disturbing host-associated microbiota, raises
con-cerns about human and environmental health
(Adamovsky et al.2018; Trevelline et al.2019). Of emerging concern are metal-based nanopar-ticles that appear as new antimicrobial agents on the market (Seil and Webster 2012). Examples of antimicrobial nanoparticles include silver, zinc oxide, titanium dioxide, copper, and iron oxide particles. These metal nanoparticles can disrupt and damage cellular membranes, DNA, and proteins, either as a result of their physical interaction with these cellu-lar components, or by inducing the formation of reactive oxygen species (Bondarenko et al. 2013; Brandelli, Ritter, and Veras2017). Additionally, metal
CONTACT Bregje W. Brinkmann b.w.brinkmann@cml.leidenuniv.nl Institute of Environmental Sciences (CML), Leiden University, Leiden, the Netherlands
Supplemental data for this article can be accessedhere.
This article has been republished with minor changes. These changes do not impact the academic content of the article.
ß 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License ( http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way.
nanoparticles release toxic metal ions, either in- or outside of cells, that exert similar adverse effects (Seil and Webster 2012; Brandelli, Ritter, and Veras 2017). Notably, nanoparticles and their shed ions do not only affect microbial cells, but can also exert adverse effects on plants and animals (Yang, Cao, and Rui2017; Sukhanova et al. 2018).
While at risk of the negative consequences of dysbiosis, some microbes can interact with anti-microbial nanoparticles, potentially reducing the nanoparticles’ toxicity to the host. In vitro studies, for instance, have demonstrated that several bac-teria can reduce toxic silver ions back into their less toxic particulate form (Lin, Lok, and Che 2014). Moreover, experiments in microcosm and meso-cosm setups revealed that microbiota can enhance their production of extracellular polysaccharides in response to chronic nanoparticle exposure (Eduok and Coulon 2017). By trapping antimicrobial nano-particles, extracellular polysaccharides presumably offer protection against toxic nanoparticles. Whether such interactions occur among host-associated microbiota in vivo, and whether these interactions significantly affect the toxicity of nano-particles to the host, is still unknown. Nevertheless, human gut microbiota have already been found to affect the toxicity of other environmental pollutants, either bioactivating or detoxifying compounds such as (nitro-)polycylic aromatic hydrocarbons, nitroto-luenes, polychlorobiphenyls, metals, and benzene derivatives (Claus, Guillou, and Ellero-Simatos2017).
In recent years, zebrafish larvae have proven to be a useful model organism to study host –micro-biota interactions in vivo (Rawls, Samuel, and Gordon 2004; Meijer, van der Vaart, and Spaink 2014). Zebrafish larvae also continue to be an important model organism in toxicology for both
human and environmental hazard assessment
(Bambino and Chu 2017; Horzmann and Freeman
2018). Similar to embryos of other teleost fish spe-cies, zebrafish embryos are assumed to develop in a sterile environment inside of the chorion, until they hatch at 2 days post-fertilization (dpf). Then, microbes that densely colonize the outer surface of chorions, and microbes from the surrounding water, likely colonize zebrafish larvae externally. Quickly thereafter, zebrafish open their mouth (at 3 dpf) and start feeding (at 5 dpf), allowing microbial col-onization of their gastrointestinal tracts (Llewellyn
et al.2014). Based on this colonization cycle, Rawls, Samuel, and Gordon (2004) established gnotobiotic techniques that enable quick and easy derivation of zebrafish larvae that are either germ-free or colon-ized by specific microbes or microbiota.
In this study, we combined gnotobiotic techni-ques for zebrafish larvae with standardized toxicity tests (Fish Embryo Acute Toxicity Test, OECD Test No. 236) enabling to explore the impact of host-associated microbiota on the acute toxicity of silver and zinc oxide nanoparticles (nAg and nZnO). Specifically, we investigated (1) how colonizing microbiota affect the sensitivity of zebrafish larvae to nAg and nZnO; (2) to what extent these impacts of microbiota–host interactions relate to the par-ticle-specific toxicity of nAg and nZnO, rather than to the toxicity of their shed Agþ and Zn2þions; and (3) how nAg and nZnO affect the abundance and the composition of colonizing microbiota. To this end, we compared the acute toxicity of nAg and nZnO between germ-free and microbially colonized zebrafish larvae. Using response addition modeling, we derived the relative contribution of ticles and their shed ions to the toxicity of nanopar-ticle suspensions. At the end of the exposures, we isolated bacteria from zebrafish larvae, and counted their abundance as an estimation of microbiota quantity. Finally, we identified the isolated colony-forming units (CFUs) based on 16S rRNA gene sequencing, to reveal what bacterial species associ-ating with zebrafish larvae are potentially resilient to nanoparticle toxicity.
2. Material and methods
2.1. Nanoparticle dispersions
Silver nanoparticles (nAg) with a primary particle size of 15 nm (NM-300K) (Klein et al. 2011) were kindly provided by RAS AG (Regensburg, Germany). These particles are commercially available as an aqueous suspension (agpureVR
W10) comprising 10% (w/w) Ag nanoparticles, 4% ammonium nitrate, 4% (w/w) polyoxyethylene glycerol trioleate, and 4%
(w/w) polyoxyethylene sorbitan mono-laurat.
Immediately prior to exposure, stock suspensions of both nanoparticles were prepared in egg water (60 mgL1 Instant Ocean sea salts; Sera GmbH, Heinsberg, Germany) at a final concentration of 100 mgL1. According to the batch dispersion protocol of the EAHC NANOGENOTOX project (v.1)
(Jensen 2018b), nAg was handled in an argon
atmosphere to prevent particle oxidation. Stock sus-pensions were stabilized by sonication for 10 min in
an ultrasonic water bath (USC200T; VWR,
Amsterdam, The Netherlands). The acoustic power of the sonicator was 12 W, as determined following the sonicator calibration standard operation proced-ure delivered in the EU FP7 NANoREG project (v. 1.1) (Jensen et al. 2018). The stock solutions were diluted to the appropriate test concentrations in egg water.
The size and morphology of both nanoparticles were characterized by transmission electron micros-copy. To this end, dispersions of 10 mgL1 nAg and nZnO were prepared in egg water as described above. FivelL of these dispersions were transferred onto 200 mesh carbon-coated copper transmission electron microscopy grids (Ted Pella, Redding, California). The grids were dried at room tempera-ture in the dark for at least 24 h. Particles on the grids were imaged with a 100 kV JEOL (Tokyo, Japan) 1010 transmission electron microscope at 50 k–60 k times magnification. The size of 50 particles from TEM images of nAg and nZnO was measured
using ImageJ software (v. 1.51 h) (Abramoff,
Magalhae, and Ram2004).
The hydrodynamic size and zeta potential of nAg and nZnO aggregates were determined using a Zetasizer Ultra instrument (Malvern Panalytical, Malvern, United Kingdom) following 0, 2, 4, 6, and 24 h of exposure (paragraph 2.3). We applied the standard operation procedure (SOP) delivered in NANoREG (v. 1.1) (Jensen 2018a), but used a fixed number of 10 runs and 3 repeated measurements per sample (n¼ 3). We selected the Smoluchowski formula for approximation of zeta potentials from electrophoretic mobility. For nAg, the refractive index (Ri) and absorption value (Rabs) were set to
0.180 and 0.010, respectively, in accordance with Bove et al. (2017). For nZnO, Ri, and Rabs were set
to 2.02 and 0.40, respectively, following the afore-mentioned SOP. Exposure concentrations below 1.5 mg nAgL1, and below 10 mg nZnOL1 are
omitted, as high variation between repeated meas-urements (SEM > 30% of the mean hydrodynamic size) and high polydispersity indices (0.70), indi-cate that the concentration of aggregates in these samples was too low for accurate dynamic light scattering analyses.
2.2. Zebrafish larvae and colonizing microbiota Embryos and larvae of ABxTL wild-type zebrafish were used for all experiments. Adult zebrafish were kept at 28C in a 14 h: 10 h light-dark cycle. Zebrafish husbandry and handling were in compli-ance with local and European animal welfare regu-lations (EU Animal Protection Directive 2010/63/EU), as surveyed by the Animal Welfare Body of Leiden University. Standard protocols (http://zfin.org) were used for the maintenance and handling of zebrafish adults and their larvae.
We divided fertilized embryos over two groups: 1. Embryos of the first group were raised
accord-ing to standard protocols (http://zfin.org). It is assumed that larvae of this group are colonized by microbes from the surrounding water and
from chorions, directly upon hatching
(Llewellyn et al.2014).
2. Embryos of the second group were sterilized and raised in autoclaved egg water, in order to exclude any microbial colonization. We steri-lized these embryos using the ‘Natural breeding method’ described by Pham et al. (2008), with the adaptations made by Koch et al. (2018). We further adapted the protocol by Koch et al. (2018) by using half of the concentration of sodium hypochlorite recommended, to ensure that all embryos hatched naturally. Briefly, embryos were incubated from 0-6 hours post fertilization (hpf) in antibiotic- and antimycotic-containing egg water (100lgmL1 Ampicillin,
5lgmL1 Kanamycin, 250 ngmL1
PA) in egg water for 5 min. Embryos were rinsed once with 10 mL sterile egg water in between these washing steps, and were rinsed thrice with 10 mL sterile egg water following both sodium hypochlorite washing steps. Only if we could not isolate any bacterial colonies from the resulting larvae on solid LB growth medium, as described in paragraph 2.5, larvae were included in this germ-free group.
As a control for the sterilization treatment, embryos of a third group were first sterilized as described for group 2, and were recolonized imme-diately thereafter by placing the embryos in egg water of the nonsterilized group 1. However, in agreement with the principles of ecological succes-sion (Odum 1969), specifically microbes with high growth rates, such as Pseudomonas aeruginosa, appeared to recolonize zebrafish embryos following initial sterilization, in favor of microbes with slower growth rates, such as Phyllobacterium
myrsinacea-rum and Sphingomonas leidyi (Supporting
Information Figure S1). For this reason, we contin-ued our experiments with embryos of groups 1 and 2 only. The embryos of both groups were incubated at 28C in petri dishes with 30 mL egg water until the start of exposure.
2.3. Exposures
Microbially colonized and germ-free zebrafish larvae were exposed to nanoparticle dispersions from 3 to 5 dpf in 24-well plates as described by Van Pomeren et al. (2017). This setup is based on OECD guideline No. 236 (OECD 2013), with the modifica-tion of exposing 10 larvae for each test concentra-tion together in one well, instead of exposing 20 larvae for each test concentration in separate wells. This modification reduces the total amount nano-material that is required per test, and produces similarly robust data to the original test (Van Pomeren et al. 2017). Three biological replicates were tested for each nominal test concentration. These were 0, 0.25, 0.75, 1, 1.5, and 2.5 mg nAgL1,
and 0, 2.5, 5, 8, 10, and 20 mg nZnOL1.
Additionally, to test the impacts of potentially shed ions, zebrafish larvae were exposed to solutions of AgNO3 and Zn(NO3)2 in egg water. The nominal
test concentrations to derive dose-response curves
for these salt solutions were 0, 0.025, 0.05, 0.1, 0.2, and 0.4 mg AgþL1; and 0, 2.5, 5, 6, 7.5, and 15 mg Zn2þL1. Because Agþ and Zn2þ ions exert anti-microbial activity, the AgNO3 and Zn(NO3)2 stock solutions can be expected to be sterile. However, since exposure of germ-free zebrafish larvae to
microbes can induce a major transcriptional
response, resulting in altered leukocyte infiltration in the intestines (Koch et al. 2018), AgNO3 and
Zn(NO3)2 stock solutions were autoclaved for this group out of precaution. Control groups were exposed to egg water without nanoparticles or cor-responding salt solutions. Exposure took place in the dark at 28C. After 24 h of exposure, dead embryos were removed, and nanoparticles and salt solutions were refreshed. Mortality was scored fol-lowing 24 h and 48 h of exposure.
The above setup is based on two assumptions, which we tested. First, we assumed that mortality in nAg exposures resulted either directly from the particles, or indirectly from their shed ions, but was not caused by the dispersion medium itself. In order to verify this assumption, particles from a 100 mgL1 stock dispersion of nAg were spun down thrice at 20 000 g for 30 min, and zebrafish larvae were exposed to the autoclaved supernatant following the above setup. Second, we assumed that the iodine and sodium hypochlorite rinsing steps that are part of the sterilization protocol, do not alter the dissolution of particles. To test this assumption, we compared particle dissolution in microbially colonized and germ-free exposures as described in paragraph 2.4.
2.4. Derivation of particle-specific toxicity
Actual concentrations of nanoparticles and their shed ions were determined using atomic adsorption spectrometry. At 0 h and 24 h following the start of exposure, 3–5 mL of the nanoparticle dispersions at each test concentration were sampled to determine total metal concentrations (n¼ 3). Another 4 mL of each nanoparticle dispersion was centrifuged for 30 min at 20 000 g, and 3 mL of the supernatant was sampled to determine metal ion concentrations (n¼ 3). The samples were acidified with 0.5% HCl and 1% HNO3, and were stored in the dark until
an Analyst 100 flame atomic absorption spectrom-eter (Perkin Elmer, Waltham, MA, USA). Elemental particle concentrations were calculated by subtract-ing ion metal concentrations from total metal con-centrations. In a few cases for nZnO, where particle concentrations were below the detection limit (as indicated by< D.L.), this calculation produced nega-tive values, which we set to zero. Particulate and total ZnO concentrations were derived from the elemental Zn concentrations based on differences in molar mass. Subsequently, replicate measure-ments were averaged, and the time weighted aver-age concentration (CTWA) was calculated for ions,
particles and total metals, as proposed for nanosaf-ety research by Zhai et al. (2016):
CTWA¼
ct¼0 hþ ct¼24 h
2 (1)
where ct¼0 h is the average concentration at 0 h, and ct¼24 h is the average concentration at 24 h
fol-lowing the start of exposure.
Finally, the mean particle-specific contribution to mortality (Eparticle) was determined for each
nano-particle test concentration by way of response add-ition (Bliss1939):
Etotal ¼ 1 ½ 1Eion
1Eparticle
(2)
where Etotal corresponds to the mean mortality in
nanoparticle exposures at the total CTWA, and Eion
corresponds to the mean mortality in AgNO3 and
Zn(NO3)2 exposures at the ion CTWA. The standard
deviation of Eparticle was derived by propagating the
standard deviation of Etotal following:
rEparticle ¼ Eparticle ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rEtotal Etotal 2 s (3) whererE
particle andrEtotal represent the standard devi-ations of Eparticle and Etotal.
2.5. Microbiota CFUs
Microbiota were isolated from zebrafish larvae at 5 dpf using a tissue homogenizer. To this end, 3 lar-vae were transferred to a 1.5 mL SafeLock microcen-trifuge tube (Eppendorf, Nijmegen, the Netherlands) comprising 200lL autoclaved egg water and 6
zir-conium oxide beads (1.0 mm-diameter; Next
Advance, New York, NY, USA). The larvae were anes-thetized for 2 min on ice, homogenized for 15 s in a tissue homogenizer (Bullet Blender model Blue-CE;
Next Advance) at speed 7, and cooled for 10 s on ice immediately thereafter. The homogenization and cooling steps were repeated 7 times to obtain a total homogenization time of 2 min.
As a measure of microbiota abundance, we determined the number of CFUs associated with lar-vae from the lowest exposure concentrations and controls at the end of exposures. Isolated micro-biota were diluted in autoclaved egg water (10, 100, and 1000 times) to reach appropriate CFU den-sities, and 100lL of the diluted microbiota was plated on LB agarose (100lL). Undiluted isolates from germ-free larvae were also plated. Following 2 days of incubation at 28C, CFUs were counted. We continued the incubation at 28C for 3 add-itional days, to check if any new colonies appeared. If colonies appeared in the germ-free group, data from the corresponding larvae were excluded from the experiment. Dilutions with the highest count-able number of CFUs below 200 were used to esti-mate microbiota abundances. It should be noted that we used our CFU estimates as a relative rather than absolute measure of microbiota abundance. Many bacteria can still not be cultured, and will thus not grow on LB growth medium. Moreover, we showed that our isolation method is detrimental to a small fraction of the isolated bacteria as pre-sented in theSupporting Information Figure S2.
Thirty colonies of nAg-exposed larvae and their controls were selected for 16S rRNA-based bacterial identification (60 colonies in total). Individual colo-nies were freshly grown on solid LB growth medium overnight at 28C, and a swap of each col-ony was lysed for 3 min in 100lL nuclease free water at 100C. Of these, a 1505-nt fragment of the 16S rRNA gene was amplified in polymerase chain
reactions (PCR) with 27 F (50-AGAGTTTGATCM
TGGCTCAG-30) and 1492 R (50-TACGGYTACCTTGTTA CGACTT-30) universal bacterial primers (Lane 1991). The PCR reactions had a total volume of 50lL and contained 1lL colony lysate, 5 lL 10 PCR buffer (200 mM Tris-HCl pH 8.4, 500 mM KCl), 5lL dNTP mix (2 mM), 1lL MgCl2 (50 mM), 0.5lL of each
pri-mer (100lM), and 0.5 lL Taq DNA polymerase
of 10 min at 72C. The DNA sequence of PCR prod-ucts was determined by BaseClear, Leiden by way
of Sanger sequencing with 27 F primers. We
trimmed low-quality areas of the obtained
sequence chromatograms, and corrected chromato-grams manually where necessary using 4Peaks soft-ware (by A. Griekspoor and Tom Groothuis; v. 1.8; nucleobytes.com). For each of the resulting sequen-ces, we performed a BLASTn search against NCBI’s nucleotide database (https://blast.ncbi.nlm.nih.gov/) to identify the corresponding species.
2.6. Statistical analyses
All statistical analyses were performed in R (v. 3.4.0; www.r-project.org). Results are reported as mean-± standard error of the mean (SEM), calculated using the‘bear’ package (v. 2.8.3; pkpd.kmu.edu.tw/ bear). All figures were plotted using Python (v. 3.6.5) with the ‘numpy’ (v. 1.15.0), ‘matplotlib’ (v. 2.2.2) and‘pandas’ (v. 0.23.3) packages.
To investigate particle dissolution, mean nano-particle concentrations at 0 h and 24 h following exposure were compared for each of the five expos-ure concentrations in a two-way ANOVA design without interaction between exposure concentra-tion and exposure time. The mean concentraconcentra-tions of shed ions at 0 h and 24 h were compared in a similar model. Subsequently, to test if the steriliza-tion procedure affected nAg dissolusteriliza-tion, we com-pared mean nAg and Agþ concentrations between exposure wells with microbially colonized and germ-free larvae using a Welch Two Sample t-test (for nAg) and Two Sample t-test (for Agþ), respect-ively. Diagnostic plots were inspected to verify if the model assumptions were met. Additionally, the Shapiro–Wilk test for normality was performed to check if residuals of the ANOVA and t-tests followed a normal distribution. We performed an F test to compare two variances to check if the variance was equally distributed over the microbially colonized and germ-free groups.
For dose-response analyses, mortality data were fitted to a three-parameter log-logistic model using the drm function of the ‘drc’ package (v. 3.0-1) (Ritz et al. 2015). The lower limit of the models was set to 0, and slope, inflection point (LC50) and upper
limit were estimated. LC50 estimates were compared
between colonized and germ-free larvae using the
compParm function. We obtained mortality esti-mates (mean and SEM) from ion (Agþ/Zn2þ) and nanoparticle (nAg/nZnO) dose-response curves, at the measured ion CTWA and total CTWA, respectively, by interpolation using the predict function. From these mortality estimates, we derived particle-spe-cific mortality estimates (mean and SEM) by way of response addition as described before (paragraph 2.4). We used these mean and SEM particle-specific mortality estimates to simulate particle-specific mor-tality data at each of the exposure concentrations (n¼ 3) using the rnorm function of the stats pack-age (v. 3.5.1). Finally, we fitted a three-parameter log-logistic function to these particle-specific mor-tality data and particle CTWA estimates, and com-pared particle-specific LC50 estimates of germ-free
and colonized larvae using the compParm function. The CFU counts of control larvae, and larvae that were exposed to the lowest exposure concentra-tions of nAg, Agþ, nZnO, and Zn2þ, were compared using an ANOVA test, combined with Tukey’s HSD post hoc test. For this model, log(xþ 1) transform-ation of CFU counts was required to ensure that the residuals of the model followed a normal
distri-bution, as indicated by the Q–Q plot and
Shapiro–Wilk test for normality. We used the diag-nostic plots of the model to check for equal vari-ance of residuals across larvae of the control, nAg, Agþ, nZnO, and Zn2þ exposures.
3. Results
3.1. Nanoparticle size, shape, aggregation, and dissolution
–20 mV over 24 h of incubation, indicating that par-ticles remained stable over the incubation time (Supporting Information Figure S3(b)).
Primary particles of nZnO had irregular shapes, with a width ranging from 12 nm to 109 nm (average 47 nm; n¼ 50), and a length ranging from 17 nm up to 234 nm (average 94 nm; n¼ 50;Figure 1). At nom-inal exposure concentrations of 10 and 20 mg nZnOL1, these primary particles formed aggregates with hydrodynamic sizes of 1086 ± 326 nm and
822 ± 193 nm at the start of incubation, and
806 ± 176 nm and 423 ± 73 nm at 24 h of incubation, respectively (Supporting Information Figure S3(c)). The corresponding zeta potential measurements indicated that nZnO aggregates stabilized over the first 2 h of incubation, reaching a zeta potential of
around –30 mV (Supporting Information
Figure S3(d)).
Immediately following dispersion, nAg and
nZnO released ions into the exposure medium (Figure 2(a,c)). No ions could be detected in con-trols without particles. Following 24 h of incubation, mean concentrations of Agþ ions in the exposure medium were still similar to those at the start of exposures (Figure 2(a,b); Supporting Information Table S1; F1,24¼ 0.025, p > .05). We note that this result needs to be interpreted with caution, as the assumption of normally distributed model residuals was not met, even following log or rank transform-ation. In accordance with the similar concentrations of Agþions measured at 0 h and 24 h of incubation, we could not detect any differences between mean mass-based particle concentrations at 0 h and 24 h of incubation for nAg (Figure 2(a,b); Supporting Information Table S1; F1,24 ¼1.1, p> .05).
Furthermore, the sterilization procedure to obtain germ-free larvae, including rinsing steps with
sodium hypochlorite and PVP-iodine, did not result in higher concentrations of Agþ, or lower
concen-trations of nAg, in the exposure medium
(Supporting Information Table S2). In contrast to Agþ, mean concentrations of Zn2þ were signifi-cantly higher following 24 h of incubation than at the start of exposures (Figure 2(c,d); Supporting Information Table S1; F1,24¼ 26.9, p ¼ 2.6 105).
The dissolution of nZnO appeared to be concentra-tion-dependent, where the release of Zn2þ seemed to have saturated already at the start of exposure at nominal concentrations below 8 mg ZnOL1, whereas concentrations of Zn2þ in the exposure medium increased over 24 h of exposure at nominal concentrations above 8 mg ZnOL1. Despite this release of ions, we did not detect differences between mass-based nZnO concentrations between 0 h and 24 h of incubation in the exposure medium (Figure 2(c,d); Supporting Information Table S1; F1,24¼ 0.26, p > .05).
3.2. Impact of microbiota on nanoparticle toxicity Zebrafish larvae that were colonized by microbes responded differently to dispersions of nAg than germ-free zebrafish larvae (Figure 3(a)). Following 48 h of exposure, median lethal toxic concentrations (LC50) were significantly higher for microbially
colonized larvae (LC50¼ 0.94 ± 0.14 mg AgL1), than for germ-free larvae (LC50¼ 0.34 ± 0.06 mg AgL1;
p¼ .0006; Figure 3(a)). Mortality among microbially colonized larvae increased from 24 h to 48 h of exposure. In contrast, nearly all mortality among germ-free larvae occurred within the first day of exposure. When nAg was removed from the disper-sion medium by centrifugation prior to exposure, and larvae were exposed for 48 h to the
dispersion medium without particles, we did not observe any mortality among larvae of each micro-biota group.
In contrast to the effects of nAg, median lethal toxic concentrations of Agþ did not differ between the microbially colonized larvae (0.14 ± 0.02 mg AgþL1) and germ-free larvae (0.16 ± 0.06 mg AgþL1; p¼ .74; Figure 3(b)) following two days of exposure. However, similar to the results of nAg exposures, nearly all mortality among germ-free lar-vae occurred during the first day of exposure to Agþ, while mortality among colonized larvae grad-ually increased over the two days of exposure.
No differences in median lethal concentrations of nZnO dispersions were observed between germ-free larvae (4.91 ± 0.43 mg ZnOL1) and colonized larvae (4.68 ± 0.62 mg ZnOL1; p> .05;
Figure 3(c)) following two days of exposure.
Moreover, mortality among both germ-free and colonized larvae increased from 24 h to 48 h of exposure to nZnO.
Similar to nZnO, we did not detect differences between median lethal concentrations of Zn2þ for colonized larvae (7.54 ± 0.82 mg Zn2þL–1) and germ-free larvae (5.68 ± 0.47 mg Zn2þL–1; p> .05; Figure 3(d)). Furthermore, mortality among Zn2þ -exposed larvae increased over the second day of exposure, independent of microbial colonization.
To explore the particle-specific contributions to the observed toxicity of nAg and nZnO, dose-response curves were corrected for the effects of shed ions in the exposure medium by way of response addition (Figure 4). The particle-specific LC50 estimates that were obtained for nAg in this way, still differed significantly between colonized (0.84 ± 0.06 mg particulate AgL–1) and germ-free lar-vae (0.34 ± 0.20 mg particulate AgL–1; t¼ 2.35, df¼ 4, p ¼ .03). At median lethal concentrations of total silver, 93 ± 13% and 42 ± 57% of the mean mortality under colonized and germ-free conditions, respectively, could be explained by the spe-cific contribution to toxicity. For nZnO,
particle-(a) (b)
(c) (d)
Figure 2. Dissolution of nAg (a,b) and nZnO (c,d) nanoparticles. Bars depict the mean concentrations of particles (black bars, left
specific median lethal toxic concentrations did not differ between microbially colonized conditions (1.94 ± 0.25 mg particulate ZnOL–1) and germ-free conditions (2.04 ± 0.48 mg particulate ZnOL–1; t¼ 0.17, df ¼ 4, p > .05). Despite the quick dissolution of these particles, the relative contribution of nZnO particles accounted 97 ± 396% and 88 ± 431% of the mean total observed mortality at median lethal con-centrations for colonized and germ-free larvae, respectively.
3.3. Impact of nanoparticles on microbiota
To investigate the impacts of nAg and nZnO on zebrafish microbiota, we isolated CFUs from zebra-fish larvae of the colonized group. At the end of the exposure time, 8.4 103 ± 3.6 103 CFUs per larvae could be isolated from the control group
(Figure 5). Exposure to the lowest test
concentrations of nAg (0.25 mgL–1), Agþ
(0.025 mgL–1), nZnO (2.5 mgL–1), and Zn2þ (2.5 mgL–1) affected this CFU count (F4,10 ¼ 45.5, p¼ 2.2 106; Figure 5). Fewer CFUs could be iso-lated from larvae that were exposed to nAg (0.89 ± 0.59 CFUs per larvae; p¼ .00002) and Agþ (1.3 102 ± 1.1 102 CFUs per larvae; p¼ .001). Exposure to nZnO or Zn2þdid not result in different CFU counts per larvae, as compared to control lar-vae (p> .05).
Considering the bactericidal effects of nAg, we further explored what bacterial species remained among the isolated CFUs following exposure to nAg, selecting 30 CFU isolates of nAg-exposed lar-vae and 30 CFU isolates of control larlar-vae for 16S rRNA gene-based identification. In total, we identi-fied 52 of 60 selected bacteria with>98% sequence identity (Supporting Information Table S3). The other 8 CFUs had low sequence quality, resulting in
(a) (b)
(c) (d)
Figure 3. Dose-response curves of microbially colonized (black markers) and germ-free (white markers) zebrafish larvae exposed
16S rRNA identity <98%. Nevertheless, BLAST results suggested similar bacterial species for these CFUs, as for the CFUs with>98% sequence identity. Hence, these records were included in the sequence identities presented below.
Based on 16S rRNA sequence identity, we identi-fied six different bacterial species among isolated CFUs (Figure 6). Additionally, we identified three groups of bacteria that we could not distinguish based on 16S rRNA sequences. Most of the isolated CFUs corresponded to Phyllobacterium myrsinacea-rum (30%), followed by bacteria of the genus Pseudomonas (30%; 13% of which was P. aerugi-nosa), Delftia lacustris/D. tsuruhatensis (17%), Rhizobium rhizoryzae (17%), and Sphingomonas lei-dyi (7%). Exposure to nAg changed the relative abundance of these bacteria among isolated CFUs.
The relative abundance of P. myrsinacearum was higher (63%) among CFUs of exposed larvae com-pared to nonexposed larvae. Additionally, we identi-fied several bacterial species that did not appear among selected CFUs of nonexposed larvae,
includ-ing Bosea sp. (13%), bacteria of the genus
Microbacterium (17%), and Staphylococcus bacteria/ Sulfitobacter donghicola (7%).
4. Discussion
Multicellular organisms live in association with diverse microbiota that contribute to host health and development. The emergence of metal-based nanoparticles on the market poses a threat to these host-associated microbiota, owing to the inherent
antimicrobial properties of these particles.
Ultimately, nanoparticle-induced perturbation of
host-associated microbiota might affect both
human and environment health (Adamovsky et al. 2018; Trevelline et al. 2019). For this reason, we set out to explore the role of zebrafish larvae-associ-ated microbiota in the acute toxicity of the two commonly applied antimicrobial nanoparticles nAg and nZnO, explicitly quantifying the relative contri-butions of particles and shed ions to toxicity.
4.1. Protection of host-associated microbiota against nanoparticles
By combining standardized acute toxicity tests and established gnotobiotic techniques, we found that colonizing microbiota protect zebrafish larvae against particle-specific lethal effects of nAg,
(a) (b)
Figure 4. Particle-specific dose-response curves of microbially colonized (black markers) and germ-free (white markers) zebrafish
larvae exposed to nAg (a), and nZnO (b) following 48 h of exposure. Error bars depict the standard error of the mean (n¼ 3).
Figure 5. Number of colony-forming units (CFUs) associated
increasing the LC50 from 0.34 ± 0.20 mg particulate
AgL–1 under germ-free conditions to 0.84 ± 0.06 mg particulate AgL–1 under microbially colonized con-ditions. Following two-days of exposure, we did not detect this microbially mediated protection against Agþ, observing similar LC50 values for microbially
colonized and germ-free larvae. This suggests that interactions between microbes and particles, rather than interactions between microbes and particle-shed ions, underlie the protective effect of micro-biota. We also did not observe any differences between the sensitivity of germ-free and colonized larvae to nZnO and Zn2þ. This similar sensitivity of germ-free and colonized larvae to nZnO and Zn2þ indicates that the protective effect against nAg results from specific interactions between microbes and particles, rather than general differences in health between germ-free and colonized larvae.
It is still unclear what mechanisms underlie the
microbially mediated protection against nAg.
Notably, the majority of zebrafish larvae that died from nAg exposure under germ-free conditions, already died within the first day of exposure, whereas mortality under microbially colonized con-ditions gradually increased over the two days of exposure. Given this acute mortality under germ-free conditions, it is possible that nAg induces an
intense pro-inflammatory immune response in
zebrafish larvae, which results in increased acute
mortality under germ-free conditions. Diverse metal nanoparticles, including silver, copper, and gold nanoparticles, have already been found to induce an acute immune response in zebrafish larvae (Brun et al. 2018; Poon et al. 2019; Van Pomeren et al. 2019). Moreover, Poon et al. (2019) found that nAg,
but not nZnO, induces inflammatory immune
responses in THP-1 cells. In case immune responses underly the differences in sensitivity between germ-free and colonized conditions, the absence of immune responses in response to nZnO might explain why we did not observe differences in sen-sitivity between germ-free and colonized larvae to nZnO. Interestingly, Koch et al. (2018) have shown that colonizing microbiota can suppress immune responses in zebrafish larvae via Myd88 signaling. Combined, these findings suggest that colonizing microbiota could protect zebrafish larvae against nAg, by suppressing pro-inflammatory immune responses that are induced by these particles.
4.2. Effects of nanoparticles on host-associated microbiota
Concurrent to the mortality among zebrafish larvae, nAg and Agþkilled the majority of zebrafish larvae-associated microbes, with barely any culturable microbes remaining after two days of exposure to nAg. Similarly high bactericidal activity of nAg has
Figure 6. Impacts of nAg (0.25 mgL–1) on the composition of CFUs isolated from zebrafish larvae (30 colonies each). Radial axes
depict log10-transformed relative abundances (%) of bacterial species. The corresponding BLAST results are included inSupporting
been demonstrated in vitro, with 4-h EC50 values
based on growth inhibition ranging from 0.35 to 18.7 mg AgL–1 for gram-negative bacteria (includ-ing several Pseudomonas species, Bacillus subtilis, and Escherichia coli), and 46.1 mg AgL–1 for the gram-positive bacterium Staphylococcus aureus (Bondarenko et al. 2013). In contrast, exposure of zebrafish larvae to sublethal concentrations of nZnO and Zn2þdid not result in a lower abundance of isolated microbes. Accordingly, in vitro studies have shown that zinc oxide particularly exhibits antimicrobial activity against Gram-positive bacteria, while the majority our isolates from zebrafish larvae were Gram-negative bacteria (Seil and Webster 2012). Moreover, the lowest concentrations of nZnO that reduced viability of the Gram-positive bacter-ium S. aureus and the Gram-negative bacterbacter-ium Escherichia coli, as determined in 24 h-in vitro expo-sures (> 400 mg ZnOL–1) (Nair et al. 2009), were well above the sublethal concentration of nZnO applied in our study (2.5 mg ZnOL–1). This could imply that nZnO and Zn2þ did not exert any bac-tericidal activity against the bacterial isolates of our study. Alternatively, growth of resistant bacteria
might have compensated for the loss of
affected bacteria.
Considering the bactericidal effects of nAg, we further investigated the effects of these particles on microbiota composition. Without exposure to nAg, CFU isolates included the opportunistic pathogenic bacteria P. aeruginosa, D. lacustris and/or D. tsuruha-tensis, and S. maltophilia (Preiswerk et al. 2011; Brooke 2012; Shin, Choi, and Ko 2012; Gellatly and Hancock 2013), and possibly S. epidermis (Otto 2009). Following exposure to nAg, we did not iso-late any of these species anymore from microbially colonized larvae. It is still unclear whether nAg elic-its immune responses that contribute to the loss of opportunistic bacteria. Only one of the bacterial iso-lates– P. myrsinacearum – appeared to be resistant against nAg. Surprisingly, this species was initially isolated from Ardisia leaf nodules (Kn€osel1984), and is known to be capable of nitrate reduction (Mergaert, Cnockaert, and Swings 2002). Since nitrate-reducing enzymes can reduce Agþ (Lin, Lok, and Che2014), it is tempting to hypothesize that P. myrsinacearum is resistant to nAg, and protects zebrafish larvae against nAg, by reducing Agþ ions that are released from nAg back into their less toxic
particulate form. However, considering other bacter-ial resistance mechanisms to silver compounds including nAg that have been identified in vitro (Silver 2003; Panacek et al. 2018), it remains to be determined what mechanisms drive bacterial resist-ance to nAg in vivo.
4.3. General applicability of the test approach This study, at the interface between toxicology and host–microbe interaction studies, is to the best of our knowledge the first of its kind. We investigate how microbial colonization affects the sensitivity of a vertebrate host to nanoparticle toxicity. In our experimental setup, we include germ-free condi-tions in nanoparticle toxicity tests, thereby combin-ing multiple stressors using standardized and
established techniques that do not require
advanced laboratory equipment (Pham et al. 2008;
OECD 2013). We note that this multistressor
research design can be applied to detect effects of nanoparticles at concentrations below the lowest-observed-effect concentrations in conventional tox-icity tests. More specifically, it can be used to screen for the interaction of host-associated microbiota with the toxicity of nanoparticles and other com-pounds of interest. Although, zebrafish that are raised in the laboratory harbor different microbiota as compared to zebrafish in their natural habitats, core groups of their microbiota are strikingly similar (Roeselers et al.2011). Moreover, despite differences in microbiota composition, hosts respond to their associated microbiota in conserved ways (Rawls et al.2006). This supports the use of our laboratory approach to include the role of host-associated microbiota in human and environmental toxicology. Similar opportunities have been established to derive germ-free Daphnia magna water fleas (Sison-Mangus, Mushegian, and Ebert 2015; Callens et al. 2016; Manakul et al.2017), extending these possibil-ities to include the role of invertebrate microbiota in toxicological research.
variables, and thus, remain hard to predict. Some bacteria, for instance, may gain resistance against nAg, and the effects thereof are still unknown. The
opportunistic pathogens P. aeruginosa and
Escherichia coli have already been found to be able protect themselves against nAg by producing adhe-sive proteins that enhance nanoparticle aggregation (Panacek et al. 2018). In case such opportunistic pathogens thrive following nAg-exposure in vivo, they might cause infections. Although, the com-plete understanding of the effects of long-term exposure to nAg is beyond the scope of this study, the finding in our study, of a profound impact of colonizing microbiota on silver nanoparticle toxicity, contributes to a better understanding of potential effects of antimicrobial nanoparticles on humans and the environment, and merits further experimen-tal attention.
5. Conclusions
In this study, we integrate the disciplines of host –-microbiota research and nanotoxicology. By com-bining gnotobiotic techniques with acute toxicity tests, we showed that host-associated microbiota protect zebrafish larvae against particle-specific toxic effects of silver nanoparticles. This protective effect was lost over time, possibly due to the bac-tericidal effects of silver particles killing protective microbes. Such indirect adverse effects of ticles, in addition to the direct impacts of
nanopar-ticles on the hosts, can be employed in
multistressor experimental designs that allow detecting otherwise hidden effects of nanoparticles. The results of our study may also contribute to understanding long-term toxic effects of nanopar-ticles, since chronic exposure of microbially
colon-ized organisms to low, yet bactericidal
concentrations of nanoparticles may enhance their sensitivity to nanoparticles over time. The observed protective effect of colonizing microbiota against silver nanoparticle toxicity moreover suggests that the effects of silver nanoparticles to humans and to the environment may be more severe following pre-exposure to antimicrobial agents. Hence, our results highlight the importance of taking micro-biota interactions into account in human and envir-onmental hazard assessment of silver nanoparticles.
Acknowledgments
We thank Wouter Beijk for assistance in the laboratory, Gerda Lamers for help with transmission electron micros-copy, Rudo Verweij for supervising atomic adsorption spec-trometry measurements and Yujia Zhai for supportive discussions about the project. We are grateful to RAS AG for providing silver nanoparticles. This work was supported by the project PATROLS of European Union’s Horizon 2020 research and innovation programme under Grant num-ber 760813.
Disclosure statement
No potential conflict of interest was reported by the authors.
Data availability statement
Data is available via Figshare https://doi.org/10.6084/m9.fig-share.c.4923261.
References
Abramoff, M. D., P. J. Magalhae, and S. J. Ram. 2004.“Image Processing with ImageJ.” Journal of Biophotonics International 11 (7): 36–42.
Adamovsky, O., A. N. Buerger, A. M. Wormington, N. Ector, R. J. Griffitt, J. H. Bisesi, Jr., and C. J. Martyniuk. 2018.“The Gut Microbiome and Aquatic Toxicology: An Emerging Concept for Environmental Health.” Environmental Toxicology and Chemistry 37 (11): 2758–2775. doi:10.1002/ etc.4249.
Bambino, K., and J. Chu. 2017.“Zebrafish in Toxicology and Environmental Health.” Current Topics in Developmental Biology 124: 331–367. doi:10.1016/bs.ctdb.2016.10.007. Bliss, C. I. 1939. “The Toxicity of Poisons Applied Jointly.”
Annals of Applied Biology 26 (3): 585–615. doi:10.1111/j. 1744-7348.1939.tb06990.x.
Bondarenko, O., A. Ivask, A. K€akinen, I. Kurvet, and A. Kahru. 2013.“Particle-Cell Contact Enhances Antibacterial Activity of Silver Nanoparticles.” PLoS One 8 (5): e64060.
Bove, P., M. A. Malvindi, S. S. Kote, R. Bertorelli, M. Summa, and S. Sabella. 2017.“Dissolution Test for Risk Assessment of Nanoparticles: A Pilot Study.” Nanoscale 9 (19): 6315–6326. doi:10.1039/C6NR08131B.
Brandelli, A., A. C. Ritter, and F. F. Veras. 2017.“Antimicrobial Activities of Metal Nanoparticles.” In Microbial Ecology, edited by C. Cravo-Laureau, C. Cagnon, B. Lauga, and R. Duran, 337–363. New York City: Springer.
Brooke, J. S. 2012. “Stenotrophomonas maltophilia: An Emerging Global Opportunistic Pathogen.” Clinical Microbiology Reviews 25 (1): 2–41. doi: 10.1128/CMR.00019-11.
Plants, Livestock, and Human Research.” Frontiers in Nutrition 5 (80): 1–15. doi:10.3389/fnut.2018.00080. Brun, N. R., B. E. V. Koch, M. Varela, W. J. G. M. Peijnenburg,
H. P. Spaink, and M. G. Vijver. 2018.“Nanoparticles Induce Dermal and Intestinal Innate Immune System Responses in Zebrafish Embryos.” Environmental Science: Nano 5 (4): 904–916. doi:10.1039/C8EN00002F.
Callens, M., E. Macke, K. Muylaert, P. Bossier, B. Lievens, M. Waud, and E. Decaestecker. 2016. “Food Availability Affects the Strength of Mutualistic Host-Microbiota Interactions in Daphnia Magna.” The ISME Journal 10 (4): 911–920. doi:10.1038/ismej.2015.166.
Claus, S., H. Guillou, and S. Ellero-Simatos. 2017. “The Gut Microbiota: A Major Player in the Toxicity of Environmental Pollutants?” NPJ Biofilms and Microbiomes 3: 17001. doi:10.1038/npjbiofilms.2016.3.
Eduok, S., and F. Coulon. 2017.“Engineered Nanoparticles in the Environments: Interactions with Microbial Systems and Microbial Activity.” In Microbial Ecology, edited by C. Cravo-Laureau, C. Cagnon, B. Lauga, and R. Duran, 63–109. New York City: Springer.
Gellatly, S. L., and R. E. W. Hancock. 2013.“Pseudomonas aer-uginosa: New Insights into Pathogenesis and Host Defenses.” Pathogens and Disease 67 (3): 159–173. doi:10. 1111/2049-632X.12033.
Hacquard, S., R. Garrido-Oter, A. Gonzalez, S. Spaepen, G. Ackermann, S. Lebeis, A. C. McHardy, et al. 2015. “Microbiota and Host Nutrition across Plant and Animal Kingdoms.” Cell Host & Microbe 17 (5): 603–616. doi:10. 1016/j.chom.2015.04.009.
Horzmann, K. A., and J. L. Freeman. 2018. “Making Waves: New Developments in Toxicology with the Zebrafish.” Toxicological Sciences 163 (1): 5–12. doi:10.1093/toxic/ kfy044.
Jensen, K. A. 2018a. Standard Operation Procedure for Intra-and Interlaboratory Reproducible Measurement of Hydrodynamic Size-Distribution and Dispersion Stability of Manufactured Nanomaterials Using Dynamic Light Scattering (DLS). Version 1.1. Bilthoven: Dutch National Institute for Public Health and the Environment (RIVM). Jensen, K. A. 2018b. The NANOGENOTOX Standard
Operational Procedure for Preparing Batch Dispersions for in Vitro and in Vivo Toxicological Studies. Version 1.2. Paris: French Agency for Food, Environmental and Occupational Health & Safety (ANSES).
Jensen, K. A., Y. Kembouche, K. Loeschner, and M. Correia. 2018. SOP for Probe-Sonicator Calibration of Delivered Acoustic Power and de-Agglomeration Efficiency for in Vitro and in Vivo Toxicological Testing. Version 1.1. Bilthoven: Dutch National Institute for Public Health and the Environment (RIVM).
Klein, C. L., S. Comero, B. Stahlmecke, J. Romazanov, T. A. J. Kuhlbusch, E. Van Doren, P-J. De Temmerman, et al. 2011. NM-Series of Representative Manufactured Nanomaterials: NM300 Silver, Characterisation, Stability, Homogeneity. Brussels: European Commission, EUR 24693 EN.
Kn€osel, D. H. 1984. “Genus IV. Phyllobacterium (Ex Kn€osel 1962) Nom. rev. (Phyllobacterium Kn€osel 1962, 96).” In Bergey’s Manual of Systematic Bacteriology, edited by N. R. Krieg and J. G. Holt, Vol 1, 254–256. Baltimore: Williams & Wilkins.
Koch, B. E. V., S. Yang, G. Lamers, J. Stougaard, and H. P. Spaink. 2018. “Intestinal Microbiome Adjusts the Innate Immune Setpoint during Colonization through Negative Regulation of MyD88.” Nature Communications 9 (1): 9. Accessed 30 October 2019.https://www.nature.com/articles/ s41467-018-06658-4doi:10.1038/s41467-018-06658-4. Lane, D. J. 1991.“16S/23S rRNA Sequencing.” In Nucleic Acid
Techniques in Bacterial Systematics, edited by E. Stackebrandt and M. Goodfellow, 115–175. New York, NY: Wiley.
Lin, I. W.-S., C.-N. Lok, and C.-M. Che. 2014.“Biosynthesis of Silver Nanoparticles from Silver(I) reduction by Periplasmic Nitrate Reductase c-Type Cytochrome Subunit NapC in a Silver-Resistant E. coli.” Chemical Sciences 5 (8): 3144–3150. doi:10.1039/C4SC00138A.
Llewellyn, M. S., S. Boutin, S. H. Hoseinifar, and N. Derome. 2014. “Teleost Microbiomes: The State of the Art in Their Characterization, Manipulation and Importance in Aquaculture and Fisheries.” Frontiers in Microbiology 5: 207. doi:10.3389/fmicb.2014.00207.
Manakul, P., S. Peerakietkhajorn, T. Matsuura, Y. Kato, and H. Watanabe. 2017. “Effects of Symbiotic Bacteria on Chemical Sensitivity of Daphnia magna.” Marine Environmental Research 128: 70–75. doi: 10.1016/j.maren-vres.2017.03.001.
Meijer, A. H., M. van der Vaart, and H. P. Spaink. 2014. “Real-Time Imaging and Genetic Dissection of Host-Microbe Interactions in Zebrafish.” Cellular Microbiology 16 (1): 39–49. doi:10.1111/cmi.12236.
Mergaert, J., M. C. Cnockaert, and J. Swings. 2002. “Pyllobacterium myrsinacearum (Subjective Synonym Phyllobacterium rubiacearum) Emend.” International Journal of Systematic and Evolutionary Microbiology 52 (5): 1821–1823. doi:10.1099/00207713-52-5-1821.
Nair, S., A. Sasidharan, V. V. D. Rani, D. Menon, S. Nair, K. Manzoor, and S. Raina. 2009. “Role of Size Scale of ZnO Nanoparticles and Microparticles on Toxicity toward Bacteria and Osteoblast Cancer Cells.” Journal of Materials Science: Materials in Medicine 20 (S1): S235–S241. doi:10. 1007/s10856-008-3548-5.
Odum, E. P. 1969.“The Strategy of Ecosystem Development.” Science 164 (3877): 262–270. doi:10.1126/science.164.3877. 262.
OECD. 2013. Fish Embryo Acute Toxicity (FET) Test. OECD Guidelines for the Testing of Chemicals No. 236. Paris: OECD.
Otto, M. 2009. “Staphylococcus epidermis – the ‘Accidental’ Pathogen.” Nature Reviews Microbiology 7 (8): 555–567. doi:10.1038/nrmicro2182.
Nanotechnology 13 (1): 65–71. doi: 10.1038/s41565-017-0013-y.
Pham, L. N., M. Kanther, I. Semova, and J. F. Rawls. 2008. “Methods for Generating and Colonizing Gnotobiotic Zebrafish.” Nature Protocols 3 (12): 1862–1875. doi:10. 1038/nprot.2008.186.
Poon, W.-L., J. C.-Y. Lee, K. S. Leung, H. Alenius, H. El-Nezami, and P. Karisola. 2019.“Nanosized Silver, but Not Titanium Dioxide or Zinc Oxide, Enhances Oxidative Stress and Inflammatory Response by Inducing 5-HETE Activation in THP-1 Cells.” Nanotoxicology. 27: 1–15. doi:10.1080/ 17435390.2019.1687776.
Preiswerk, B., S. Ullrich, R. Speich, G. V. Bloemberg, and M. Hombach. 2011.“Human Infection with Delftia tsuruhaten-sis Isolated from a Central Venous Catheter.” Journal of Medical Microbiology 60 (2): 246–248. doi:10.1099/jmm.0. 021238-0.
Rawls, J. F., B. S. Samuel, and J. I. Gordon. 2004.“Gnotobiotic Zebrafish Reveal Evolutionarily Conserved Responses to the Gut Microbiota.” Proceedings of the National Academy of Sciences of the United States of America 101 (13): 4596–4601. doi:10.1073/pnas.0400706101.
Rawls, J. F., M. A. Mahowald, R. E. Ley, and J. I. Gordon. 2006. “Reciprocal Gut Microbiota Transplants from Zebrafish and Mice to Germ-Free Recipients Reveal Host Habitat Selection.” Cell 127 (2): 423–433. doi:10.1016/j.cell. 2006.08.043.
Ritz, C., F. Baty, J. C. Streibig, and D. Gerhard. 2015. “Dose-Response Analysis Using R.” PLoS One 10 (12): e0146021. doi:10.1371/journal.pone.0146021.
Roeselers, G. E. K. Mittge, W. Z. Stephens, D. M. Parichy, C. M. Cavanaugh, K. Guillemin, and J. F. Rawls. 2011. “Evidence for a Core Gut Microbiota in the Zebrafish.” The ISME Journal 5 (10): 1595–1608. doi:10.1038/ismej.2011.38. Seil, J. T., and T. J. Webster. 2012.“Antimicrobial Applications of
Nanotechnology: Methods and Literature.” International Journal of Nanomedicine 7: 2767–2781. doi:10.2147/IJN.S24805. Shin, S. Y., J. Y. Choi, and K. S. Ko. 2012. “Four Cases of Possible Human Infections with Delftia lacustris.” Infection 40 (6): 709–712. doi:10.1007/s1501-012-0339-1.
Silver, S. 2003.“Bacterial Silver Resistance: Molecular Biology and Use and Misuses of Silver Compounds.” FEMS Microbiology Reviews 27 (2–3): 341–353. doi:10.1016/ S0168-6445(03)00047-0.
Singh, C., S. Friedrichs, M. Levin, R. Birkedal, K. A. Jensen, G. Pojana, W. Wohlleben, et al. 2011. Zinc oxide 110, NM-111, NM-112, NM-113: Characterisation and Test Item Preparation. Brussels: European Commission, EUR 25066 EN.
Sison-Mangus, M. P., A. A. Mushegian, and D. Ebert. 2015. “Water Fleas Require Microbiota for Survival, Growth and Reproduction.” The ISME Journal 9 (1): 59–67. doi:10.1038/ ismej.2014.116.
Sukhanova, A., S. Bozrova, P. Sokolov, M. Berestovoy, A. Karaulov, and I. Nabiev. 2018. “Dependence of Nanoparticle Toxicity on Their Physical and Chemical Properties.” Nanoscale Research Letters 13 (1): 13. doi:10. 1186/s11671-018-2457-x.
Trevelline, B. K., S. S. Fontaine, B. K. Hartup, and K. D. Kohl. 2019. “Conservation Biology Needs a Microbial Renaissance: A Call for the Consideration of Host-Associated Microbiota in Wildlife Management Practices.” Proceedings of the Royal Society B: Biological Sciences 286 (1895): 20182448. doi:10.1098/rspb.2018.2448.
Van Pomeren, M., W. J. G. M. Peijnenburg, N. R. Brun, and M. G. Vijver. 2017. “A Novel Experimental and Modeling Strategy for Nanoparticle Toxicity Testing Enabling the Use of Small Quantities.” International Journal of Environmental Research and Public Health 14 (11): 1348. doi:10.3390/ijerph14111348.
Van Pomeren, M., W. J. G. M. Peijnenburg, R. C. Vlieg, S. J. T. Van Noort, and M. G. Vijver. 2019. “The Biodistribution and Immuno-Responses of Differently Shaped Non-Modified Gold Particles in Zebrafish Embryos.” Nanotoxicology 13 (4): 558–5404. doi:10.1080/17435390. 2018.1564079.
Willing, B. P., S. L. Russel, and B. B. Finlay. 2011.“Shifting the Balance: Antibiotic Effects on Host-Microbiota Mutualism.” Nature Reviews Microbiology 9 (4): 233–243. doi:10.1038/ nrmicro2536.
Yang, J., W. Cao, and Y. Rui. 2017. “Interactions Between Nanoparticles and Plants: Phytotoxicity and Defense Mechanisms.” Journal of Plant Interactions 12 (1): 158–169. doi:10.1080/17429145.2017.1310944.
